Also honestly the *long-term* solution is questioning why we prioritize gendering in our pronouns.
Pronouns are scoped temporary variables. If you were to design a language today, would you choose to prioritize that information? We know that gendered pronouns automatically load a giant set of preconceptions, whether we want to or not. Does doing so help or hurt more often?
@civodul Though English's 'default' "they" (which I think most people do in fact have in their grammars, so this sort of use of singular "they" historically pre-dates the use of "you" (rather than "thou") as a singular) actually seems to have particular properties, which also partially overlap with the use of "they" as a chosen pronoun.
That is, English's default non-plural "they" signals either lack of knowledge or genericity. Using it to refer to known, specific person still usually takes English speakers conscious effort.
But, yeah, Romance languages in general end up in a 'worse' position in terms of their accident morphology: there are only two genders and they don't generally neutralise in the plural.
This is my favorite line of the talk, on why one would make a language:
When we write code, we are restructuring our thoughts to fit the language of the computer.
When we design our own languages, we rewire the computer to match the structure of our thoughts.
Something that @AlexKnauth, @ArneBab, and I have been slowly working on: yet another lisp syntax that removes the parentheses called Wraith: https://github.com/cwebber/racket2-rfcs/blob/wraith/wraith/wraith.md
Still in progress but it looks fairly natural I think, and preserves the better parts of Lisp.
one of my favorite features of scribble is that the racket docs search searches not just for functions and types and code stuff but also page titles, filenames, and importantly technical terms which have a specific way to define them in the markup and then you can quickly look for terms like "synchronizable event" and immediately jump to the definition of the term
which is really good!!! it makes the docs extremely usable, much more than javadocs
@ionchy Yup. Ryan Culpepper gave a talk on proper design with macros; I can't find it though. Common mistakes are (1) using macros when functions will do and (2) expanding to lots of code when you ought to expand to a function call (3) not let-binding expressions that are inputs that appear multiple times in the template, causing duplicate evaluation.
@ionchy I'm not so sure. If the language is designed to allow the user to extend it, it seems like the editor must have some way to adapt to user-defined extensions. Making that external to the language is leads to a ton of code duplication (each editor will invent it's own language for control customization, which will not be portable). Making it possible to do language-extension with IDE-extension as a first-class feature is probably a good idea.
@wilbowma well both are good... we can coexist in peace and harmony... all we need now is an avatar, master of both macros and monads
racism, spicy take on iq
history of iq tests:
-invented by racists with confirmation bias
-historically used by racists to justify racist education policies and also eugenics
-currently used by racists to justify racist education policies and also eugenics
-debunked countless times
-still taken seriously because ????????
I like programming languages, functional programming, macros, and type systems
pronouns: he/him or they/them
A Mastodon instance for programming language theorists and mathematicians. Or just anyone who wants to hang out.